Throughout your life, you are given bits of advice, by experts, by newspapers, by people who have done it themselves and know what they are talking about and by people who have no idea what they are talking about. And you filter it for whether it makes sense and whether it dovetails nicely with what you have already decided to do.

Then you get pregnant and a number of factors change: first, the advice is much more patronising, and much more militant. Take last Friday's announcement that, instead of allowing themselves the odd glass of wine, as previously recommended, from now on pregnant women should drink nothing at all. Not based on any new research, by the way, just "to be on the safe side". Second, you lose the nonchalant "if cheese really were poisonous, he would have died by now" attitude you had when you got your first puppy, say, and freely fed it treats. And third, just when you start to take advice most seriously, it conflicts. I'll give you an example - the other day, I went to the doctor with a kidney infection, and she said: "Cocodamol is the best thing to take when you're pregnant." I walked across the Asda car park into the pharmacy, where the pharmacist said: "You can't take cocodamol, you're pregnant." So I did what the pharmacist told me to, because it seemed rude to ignore the advice of the person standing right in front of me - but that's not necessarily the best way to make medical decisions, is it?

So amid the barrage of advice, what should you take seriously, and what can you ignore? What's right, and what's just scary nonsense? I decided to do some research on the subject, rather than just taking the government's word for it.

Alcohol

It is a shame, but there is very little research done on the prevailing attitudes to the drunk pregnant person, so we will have to rely on patched-together accounts from my mum and my own memory. Thirty years ago, it was not the done thing to get rolling drunk while pregnant, but it wasn't the done thing for non-pregnant women either. You could certainly drink moderately while pregnant, however. Ten years ago, an EastEnders storyline had Tiffany out on the razzle while she was up the duff, and Grant getting upset. EastEnders is a brilliant barometer for a certain sort of acceptability - it isn't an accurate portrait of anything, but it does indicate that in 1996, middle-class people definitely thought getting lashed up while with child was dangerous, and that it was probably something working-class people did. Still, a glass of wine a day was considered a decent way to proceed.

Five years ago, the government advice was as it was until last Thursday, one or two drinks once or twice a week, but it was taken as given that this was akin to government advice on regular drinking - that is, "14 units a week, are you kidding?", and nobody frowned upon the expectant mother having a drink a day.

For the past two years, however, the Daily Mail has been obsessively reporting surveys saying expectant mothers should not drink at all as there is no safe lower limit, and for the past year, this has been bread and butter for right-wing broadsheets too, plus the Observer.

A typical article about foetal alcohol spectrum disorder (FASD) will start like this: a case study will be presented, in which a child with the full range of symptoms is described. These include mental impairment, very low birthweight and subsequent growth restriction, and a distinctive phenotype (a small and misshapen head, widely spaced eyes, thin lips and a flat philtrum). Then it will describe the rapid growth in FASD, without pointing out that this correlates exactly with the rise in binge drinking. (Nobody is in any doubt that binge drinking is bad for foetal health. The question mark is over whether or not we should be abstaining completely.) Then it will say something like, in America, pregnant women are advised not to drink at all, and then a doctor - quite often it is Raja Mukherjee, consultant psychiatrist at the Surrey and Borders Partnership trust - will be quoted thus: "The uncertain level of individual risk to the developing foetus, coupled with the possibility of misinterpreting a health promotion message, mean that the only safe message in pregnancy is abstinence from alcohol." The piece, if it's in a quality newspaper, will then rather sheepishly deliver the information that the original case study was the child of a serious alcoholic, drinking upwards of a bottle of vodka a day. In a tabloid, this will be passed over.

Last week the government guidelines changed, and we are now told that abstinence is the only responsible choice, because, as the Guardian put it, "Ministers believe the change in guidance on alcohol is needed because too many women underestimate the risks to their baby, although there is no new scientific evidence."

All this despite the fact that a study in 2006 by the British Journal of Obstetrics and Gynaecology concluded that there was no convincing evidence of adverse effects of prenatal alcohol exposure at low to moderate levels, where moderate was defined as 10.5 units per week (but, the study author Dr Robert Fraser reminds us, not at one sitting, obviously).

Physiologically and sociologically, it just does not make sense that small amounts of alcohol are bad for you when pregnant. As Dr Eric Jauniaux, professor of obstetrics and foetal medicine at the Royal Free hospital in London, points out: "Alcohol is mainly metabolised by the liver, and only what's left will be met by the placenta. The amount that could reach the foetus in a glass of beer or a glass of wine is negligible. I would be much more concerned with breastfeeding and drinking." Jauniaux, incidentally, has been studying transfer through the placenta for the past 20 years, is one of the leading national experts on the matter and yet is never quoted in connection with any of the scare stories you read on alcohol and unborn babies. And sociologically, of course, Jauniaux reminds us: "How long have people been drinking wine or beer, thousands of years?"

Many of our mothers drank during pregnancy, and while this may not be very scientific, the lack of alcohol-related brain damage in our generation should militate against blank credulity when you are told not to touch a drop.

The obvious question now is, what is this enthusiasm for giving women advice that is strict to the point of being unscientific?

First, consider Mukherjee's perspective, which is not of foetal research but of neuroscience. He sees children with, for example, behavioural problems, poor concentration and hyperactivity disorder. Foetal alcohol spectrum disorder is a diagnosis of exclusion, in that once you have ruled out a raft of medical conditions whose current prevalence is largely unexplained, then you diagnose FASD. Research on the impact of ethanol on pregnant rats leads Mukherjee to believe that small levels of alcohol can be dangerous in humans, but obstetrics experts such as Robert Fraser at the National Perinatal Epidemology Unit have not been sufficiently impressed by the research to let it sway their conclusions. Anyway, Mukherjee maintains - I am paraphrasing, but not unfairly, I think - that, when you have this web of often indistinguishable conditions, ruinous to the lives of children and, of course, their parents, if there is even a possibility that alcohol could be causing any of it, then can't we at least stop these pregnant women drinking altogether so we can see what is going on?

You can see his frustration, but from the point of view of the pregnant woman, presented with no bona fide evidence for foetal harm at low drinking levels, the response is, of course, "Sod that."

Incidentally, in America, the great white hope of the teetotal pregnancy, oft cited by all the teetotal campaigners, the single greatest cause of foetal growth restriction is crack addiction. There has never been any controversy about this - crack addiction is strongly discouraged during pregnancy as it is illegal. And yet the message of abstinence has not worked.

Abstinence messages never work. Everybody knows they don't work, and I would go one further and say that social conservatives never intend them to work - they intend, rather, with their stringency, to effect a severance between the state and the individual. Don't come crying to us if it all goes wrong. We have already warned you to be perfect.

Listeria

If you have the misfortune not to be pregnant, you might not know that the expectant mother is instructed by everyone, very forcefully, not to eat blue cheese or any cheese that hasn't been pasteurised, likewise unpasteurised milk and ice cream. Also you must not eat: rinded cheese, goat's cheese, pâté, terrines and rillettes, sushi, sashimi and artisan-smoked fish (as opposed to the industrial sort, which is smoked to death). This is besides the other fish injunctions, which tend to be about mercury - oh, and bagged salad.

Basically, if you can get it in a deli and it's costed per 100g, you probably aren't allowed to eat it. Most of these injunctions are about avoiding listeriosis, which is the disease proceeding from the listeria bacteria (apart from the mercury stuff, and "no liver" injunctions, which are about an excess of vitamin A).

Listeria has been my particular bugbear ever since a midwife - that is, a trained prenatal professional who, unless I develop complications, represents the highest medical authority I can expect to deal with throughout my pregnancy - told me that I could get listeriosis, thereby brain-damaging my foetus, without knowing about it. Now, listeriosis is an incredibly serious disease, with extremely serious symptoms, taken extremely seriously by epidemiologists nationwide. Get it without noticing it? If I got listeriosis, the national papers would know about it. It would be the third outbreak that has occurred in this country in the past 20 years. My beloved, C, said: "Well, she was just erring on the side of caution." This is a common line. But the distinction between "caution" and "misinformation" could not be more important - when it is blurred, as it so often is, the upshot is that pregnant women either become neurotic, or lose faith in the medical profession altogether.

Here are some other things that are wantonly untrue: pasteurisation, in fact, has nothing to do with a cheese's ability to harbour the listeria bacteria. The bacteria that characterise different cheeses are introduced after the pasteurisation process anyway. Listeria flourishes in moist environments, so parmesan is safe where camembert isn't, but even rinded and soft cheeses are safe once they have been cooked. But food hygiene is a much more important factor than moisture - raw fish does not come out of the sea carrying listeria, but contracts the bacteria from contact with dirty hands. Of the past two outbreaks of listeria in Britain, one was from butter and the other from lettuce (there have been other instances of product recalls, but no human contamination).

In fact the three worst recorded cases of list-eria since 1992 have all been in France, and were all from pork tongue in jelly, which nobody in their right mind would ever eat. Of the past 10 listeriosis outbreaks in America, only two were from cheese, and one of those was a Mexican homemade cheese. The notion that there are pregnant people out there whipping themselves into a frenzy of guilt because they have eaten some gorgonzola is just infuriating.

Jauniaux cites an interesting difference between Google hit rates and those on PubMed, a collected source of peer-reviewed papers in medical journals. "Listeria and pregnancy - Google, 190,000 hits; PubMed, 107 hits. Cheese and listeria - Google, 194,000 hits; PubMed, 169 hits. Sushi and pregnancy - Google, 628,000 hits; PubMed, 0 hits. Raw fish and listeria: Google, 123,000 hits; PubMed, 49 hits. You can see immediately," Jauniaux concludes, "the disproportion between the epidemiological evidence and the general public hysteria about the disease."

I wrote a potted rant about listeria hysteria some weeks ago in this paper, and readers emailed me accusing me of murdering people's foetuses with misinformation. Which is completely foolish. Jauniaux gave a lecture at the annual conference of bacteriologists two years ago, and no one in attendance had come across an outbreak of listeria in years. In England and Wales, the last figures for listeriosis were 2.7 cases per million (and that is among the entire population, not pregnant women). This risk is not negligible - tolerable risk is defined by the government as one in a million - but it is pretty small and, more importantly, the advice is misleading. In fact the potential danger comes from ready-to-eat foods: it could be anything that gives you listeria, it could be cheese or butter or lettuce or coleslaw; an ice-cream cake or a rice salad. In Spain, pregnant women are told not to eat any salad. All of this is overblown, and none of it is as effective as a simple food hygiene message, telling you to observe the basic rules that you would hold to if you were cooking for people you didn't want to poison: wash knives and boards, wash foodstuffs and so on.

And by the way: the most dangerous places to eat in Britain are hospitals and old people's homes, and that is before they have even had a go at treating you.

Toxoplasmosis

Foetal exposure to the disease results in brain damage, blindness or possibly death. Famously, you get it from cat poo: except that almost nobody ever does. Jauniaux points out: "The cat will only give it to you when it has infected itself. It doesn't live in the cat."

In fact, undercooked lamb and beef (all meat could carry it, but it is not considered safe to eat pink chicken or pork) are the cause of toxoplasmosis most of the time, and you can also get it from root vegetables. In short, you can get it from anything that grows in the ground or that eats plants from the ground. Again, thorough cooking banishes the disease, so there is no great mystery to avoiding it.

Rates, however, are very high - vastly greater, at one in 2,000 mothers-to-be getting infected, than with listeria. Rates are broadly comparable to Down's syndrome, which, of course, is screened for. In France, rates are one per 500 pregnant women, but "most of the population in France is immune, which is interesting. They have a much more varied diet," Jauniaux observes. These figures are a bit complicated, since the likelihood of the mother transmitting to the foetus goes up dramatically according to gestational age (first trimester, 5% chance, final trimester, 80%), but the chances of the foetus being brain damaged by the disease decrease with its gestational age.

It is not necessarily helpful to talk about stamping the disease out altogether, since having had it pre-pregnancy (this is known as "seroprevalence", which is much greater in France but even there has gone down in the past two decades) cancels your chances of passing it on to your offspring.

Most interesting of all is this: toxoplasmosis can be screened for during pregnancy, and can also treated in the womb, which is not the case for Down's syndrome. This is your classic authoritarian sleight of hand - with a scattergun delivery of partly accurate, partly misrepresented and partly plain wrong information, all our attention is focused on our individual responsibilities, which we will probably, through a combination of waywardness and stupidity, fail to fulfil anyway. Yet, statistically, it would be a better investment for our unborn children if we were to eat whatever we liked and spent our time and energy pressuring MPs for better prenatal screening and better standards of hygiene in medical institutions.

Caffeine

Excessive consumption of caffeine has been linked to miscarriage, cleft palate and low birthweight. It is frowned on far more in the US than here. There, pregnant women are told to stop eating chocolate, despite the fact that the average chocolate bar has 30mg of caffeine, and 240mg a day has always been considered a safe amount. In other words, you would have to eat eight bars of chocolate a day to tip the scales, in which case your risk factor would more probably be that you were obese and/or suffering from malnutrition.

If a clearly pregnant woman were to order a double espresso in America, it would be on about the same scale of taboo as Kate Moss being photographed here buying a pack of fags while pregnant. Even in the UK, the advice is "three cups of coffee a day", though I can't tell you (and don't think I haven't looked) whether this means decent coffee or service-station coffee, nor whether that advice is on the basis of miscarriage-risk or another factor, nor, for that matter, how big the cups of coffee are.

At the beginning of this year, a Danish study published findings on this subject: 1,200 women were surveyed, which is more than double the sample of the next largest study. The adjustments for confounding factors were comprehensive, and the conclusion was that no significant differences occurred between the caffeine drinkers and the decaff drinkers, in the birthweight of their babies or the frequency of preterm delivery: even among women who drank more than seven cups of coffee a day.

But when a midwife - a different one from the listeria woman - said, "Don't drink coffee", and I said, "There's actually this Danish study out ...", she said: "Who do you think knows better, a doctor, or a Danish study?" As if there were something inherently untrustworthy about being Danish. But besides that, well, let's see - who would know more? A general practitioner or a research body conducting the largest European study on record?

As a pregnant woman, you are expected to relinquish your critical judgment altogether; the atmosphere seems to be, "You can make your own stupid judgments when it's about your own stupid skin, but you are in charge of a foetus now. You'll probably do it wrong."

In conclusion

Despite having got yourself into a spot of bother, pregnant women, you are still in the possession of adult judgment, and you are still allowed to use it. If the official advice sounds stringent to the point of insanity, examine it more closely - you aren't just a selfish person, looking for loopholes for your own sorry enjoyment.

Try to remember, when the advice turns out to be nonsense, that not everyone has evil motives: some people will give you bad advice because they are stupid or ill-informed. Others will give you bad advice because, without even realising it, they have a yen to bring the business of procreation under closer central control. They just don't trust you. But then, why should they? You are an absurd shape and you keep crying.